Discrete Choice Modeling William Greene Stern School of Business New York University.

Slides:



Advertisements
Similar presentations
Discrete Choice Modeling William Greene Stern School of Business IFS at UCL February 11-13, 2004
Advertisements

Discrete Choice Modeling William Greene Stern School of Business New York University.
3. Binary Choice – Inference. Hypothesis Testing in Binary Choice Models.
Discrete Choice Modeling William Greene Stern School of Business New York University.
12. Random Parameters Logit Models. Random Parameters Model Allow model parameters as well as constants to be random Allow multiple observations with.
[Part 1] 1/15 Discrete Choice Modeling Econometric Methodology Discrete Choice Modeling William Greene Stern School of Business New York University 0Introduction.
Discrete Choice Modeling William Greene Stern School of Business New York University.
Discrete Choice Models William Greene Stern School of Business New York University.
Discrete Choice Modeling William Greene Stern School of Business New York University Lab Sessions.
15. Stated Preference Experiments. Panel Data Repeated Choice Situations Typically RP/SP constructions (experimental) Accommodating “panel data” Multinomial.
Discrete Choice Modeling William Greene Stern School of Business New York University.
10. Multinomial Choice. A Microeconomics Platform Consumers Maximize Utility (!!!) Fundamental Choice Problem: Maximize U(x 1,x 2,…) subject to prices.
Discrete Choice Modeling William Greene Stern School of Business New York University.
Discrete Choice Modeling William Greene Stern School of Business New York University.
[Part 15] 1/24 Discrete Choice Modeling Aggregate Share Data - BLP Discrete Choice Modeling William Greene Stern School of Business New York University.
Discrete Choice Models William Greene Stern School of Business New York University.
Discrete Choice Modeling William Greene Stern School of Business New York University.
Discrete Choice Modeling William Greene Stern School of Business New York University.
Empirical Methods for Microeconomic Applications University of Lugano, Switzerland May 27-31, 2013 William Greene Department of Economics Stern School.
[Part 9] 1/79 Discrete Choice Modeling Modeling Heterogeneity Discrete Choice Modeling William Greene Stern School of Business New York University 0Introduction.
Discrete Choice Modeling William Greene Stern School of Business New York University Lab Sessions.
1/54: Topic 5.1 – Modeling Stated Preference Data Microeconometric Modeling William Greene Stern School of Business New York University New York NY USA.
[Part 4] 1/43 Discrete Choice Modeling Bivariate & Multivariate Probit Discrete Choice Modeling William Greene Stern School of Business New York University.
Discrete Choice Modeling William Greene Stern School of Business New York University Lab Sessions.
Part 24: Stated Choice [1/117] Econometric Analysis of Panel Data William Greene Department of Economics Stern School of Business.
Discrete Choice Modeling
Discrete Choice Modeling William Greene Stern School of Business New York University.
Discrete Choice Modeling William Greene Stern School of Business New York University.
14. Scaling and Heteroscedasticity. Using Degenerate Branches to Reveal Scaling Travel Fly Rail Air CarTrain Bus LIMB BRANCH TWIG DriveGrndPblc.
Microeconometric Modeling
1/30: Topic 4.1 – Nested Logit and Multinomial Probit Models Microeconometric Modeling William Greene Stern School of Business New York University New.
[Part 7] 1/96 Discrete Choice Modeling Multinomial Choice Models Discrete Choice Modeling William Greene Stern School of Business New York University 0Introduction.
Meeghat Habibian Analysis of Travel Choice Transportation Demand Analysis Lecture note.
Discrete Choice Modeling William Greene Stern School of Business New York University.
Discrete Choice Modeling William Greene Stern School of Business New York University.
1/53: Topic 3.1 – Models for Ordered Choices Microeconometric Modeling William Greene Stern School of Business New York University New York NY USA William.
11. Nested Logit Models. Extended Formulation of the MNL Groups of similar alternatives Compound Utility: U(Alt)=U(Alt|Branch)+U(branch) Behavioral implications.
[Part 8] 1/26 Discrete Choice Modeling Nested Logit Discrete Choice Modeling William Greene Stern School of Business New York University 0Introduction.
Discrete Choice Modeling William Greene Stern School of Business New York University.
1/30: Topic 4.1 – Nested Logit and Multinomial Probit Models Microeconometric Modeling William Greene Stern School of Business New York University New.
Discrete Choice Modeling William Greene Stern School of Business New York University.
6. Ordered Choice Models. Ordered Choices Ordered Discrete Outcomes E.g.: Taste test, credit rating, course grade, preference scale Underlying random.
1/122: Topic 3.3 – Discrete Choice; The Multinomial Logit Model Microeconometric Modeling William Greene Stern School of Business New York University New.
[Part 5] 1/43 Discrete Choice Modeling Ordered Choice Models Discrete Choice Modeling William Greene Stern School of Business New York University 0Introduction.
Discrete Choice Modeling William Greene Stern School of Business New York University.
Discrete Choice Modeling William Greene Stern School of Business New York University.
Microeconometric Modeling
William Greene Stern School of Business New York University
William Greene Stern School of Business New York University
Discrete Choice Modeling
William Greene Stern School of Business New York University
Econometric Analysis of Panel Data
Discrete Choice Models
Discrete Choice Modeling
Microeconometric Modeling
Microeconometric Modeling
Microeconometric Modeling
Econometrics Chengyuan Yin School of Mathematics.
Discrete Choice Modeling
Microeconometric Modeling
Microeconometric Modeling
Econometric Analysis of Panel Data
William Greene Stern School of Business New York University
William Greene Stern School of Business New York University
Microeconometric Modeling
Econometric Analysis of Panel Data
Empirical Methods for Microeconomic Applications University of Lugano, Switzerland May 27-31, 2019 William Greene Department of Economics Stern School.
William Greene Stern School of Business New York University
Empirical Methods for Microeconomic Applications
Presentation transcript:

Discrete Choice Modeling William Greene Stern School of Business New York University

Part 9 Multinomial Logit Models

A Microeconomics Platform  Consumers Maximize Utility (!!!)  Fundamental Choice Problem: Maximize U(x 1,x 2,…) subject to prices and budget constraints  A Crucial Result for the Classical Problem: Indirect Utility Function: V = V(p,I) Demand System of Continuous Choices  The Integrability Problem: Utility is not revealed by demands

Implications for Discrete Choice Models  Theory is silent about discrete choices  Translation of utilities to discrete choice requires: Well defined utility indexes: Completeness of rankings Rationality: Utility maximization Axioms of revealed preferences  Choice sets and consideration sets – consumers simplify choice situations  Implication for choice among a set of discrete alternatives  This allows us to build “models.” What common elements can be assumed? How can we account for heterogeneity?  However, revealed choices do not reveal utility, only rankings which are scale invariant.

Multinomial Choice Among J Alternatives Random Utility Basis U itj =  ij +  i ’x itj +  ij z it +  ijt i = 1,…,N; j = 1,…,J(i,t); t = 1,…,T(i) N individuals studied, J(i,t) alternatives in the choice set, T(i) [usually 1] choice situations examined. Maximum Utility Assumption Individual i will Choose alternative j in choice setting t iff U itj > U itk for all k  j. Underlying assumptions Smoothness of utilities Axioms: Transitive, Complete, Monotonic

Features of Utility Functions  The linearity assumption U itj =  ij +  i ’x itj +  j z it +  ijt To be relaxed later: U itj = V(x itj,z it,  i ) +  ijt  The choice set: Unordered alternatives j = 1,…,J(i,t)  Deterministic and random components  Generic vs. alternative specific components Attributes of choices, x itj and characteristics of the chooser, z it. Coefficients  Alternative specific constants  ij may vary by individual  Preference weights,  i may vary by individual  Individual components,  j typically vary by choice, not by person  Scaling parameters, σ = Var[ε], subject to much modeling

The Multinomial Logit (MNL) Model  Independent extreme value (Gumbel): F(  itj ) = 1 – Exp(-Exp(  itj )) (random part of each utility) Independence across utility functions Identical variances (means absorbed in constants) Same parameters for all individuals (temporary)  Implied probabilities for observed outcomes

Specifying the Probabilities Choice specific attributes (X) vary by choices, multiply by generic coefficients. E.g., TTME=terminal time, GC=generalized cost of travel mode Generic characteristics (Income, constants) must be interacted with choice specific constants. Estimation by maximum likelihood; d ij = 1 if person i chooses j

Using the Model to Measure Consumer Surplus

Willingness to Pay

Observed Data  Types of Data Individual choice Market shares – consumer markets Frequencies – vote counts Ranks – contests, preference rankings  Attributes and Characteristics Attributes are features of the choices such as price Characteristics are features of the chooser such as age, gender and income.  Choice Settings Cross section Repeated measurement (panel data)  Stated choice experiments  Repeated observations – THE scanner data on consumer choices

Data on Discrete Choices CHOICE ATTRIBUTES CHARACTERISTIC MODE TRAVEL INVC INVT TTME GC HINC AIR TRAIN BUS CAR AIR TRAIN BUS CAR AIR TRAIN BUS CAR AIR TRAIN BUS CAR

Estimated MNL Model Discrete choice (multinomial logit) model Dependent variable Choice Log likelihood function Estimation based on N = 210, K = 5 Information Criteria: Normalization=1/N Normalized Unnormalized AIC Fin.Smpl.AIC Bayes IC Hannan Quinn R2=1-LogL/LogL* Log-L fncn R-sqrd R2Adj Constants only Chi-squared[ 2] = Prob [ chi squared > value ] = Response data are given as ind. choices Number of obs.= 210, skipped 0 obs Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] GC| *** TTME| *** A_AIR| *** A_TRAIN| *** A_BUS| ***

Discrete choice (multinomial logit) model Dependent variable Choice Log likelihood function Estimation based on N = 210, K = 5 Information Criteria: Normalization=1/N Normalized Unnormalized AIC Fin.Smpl.AIC Bayes IC Hannan Quinn R2=1-LogL/LogL* Log-L fncn R-sqrd R2Adj Constants only Chi-squared[ 2] = Prob [ chi squared > value ] = Response data are given as ind. choices Number of obs.= 210, skipped 0 obs Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] GC| *** TTME| *** A_AIR| *** A_TRAIN| *** A_BUS| *** Estimated MNL Model

Discrete choice (multinomial logit) model Dependent variable Choice Log likelihood function Estimation based on N = 210, K = 5 Information Criteria: Normalization=1/N Normalized Unnormalized AIC Fin.Smpl.AIC Bayes IC Hannan Quinn R2=1-LogL/LogL* Log-L fncn R-sqrd R2Adj Constants only Chi-squared[ 2] = Prob [ chi squared > value ] = Response data are given as ind. choices Number of obs.= 210, skipped 0 obs Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] GC| *** TTME| *** A_AIR| *** A_TRAIN| *** A_BUS| *** Estimated MNL Model

Model Fit Based on Log Likelihood  Three sets of predicted probabilities No model: P ij = 1/J (.25) Constants only: P ij = (1/N)  i d ij [(58,63,30,59)/210=.286,.300,.143,.281) Estimated model: Logit probabilities  Compute log likelihood  Measure improvement in log likelihood with R-squared = 1 – LogL/LogL0 (“Adjusted” for number of parameters in the model.)  NOT A MEASURE OF “FIT!”

Estimated MNL Model Discrete choice (multinomial logit) model Dependent variable Choice Log likelihood function Estimation based on N = 210, K = 5 Information Criteria: Normalization=1/N Normalized Unnormalized AIC Fin.Smpl.AIC Bayes IC Hannan Quinn R2=1-LogL/LogL* Log-L fncn R-sqrd R2Adj Constants only Chi-squared[ 2] = Prob [ chi squared > value ] = Response data are given as ind. choices Number of obs.= 210, skipped 0 obs Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] GC| *** TTME| *** A_AIR| *** A_TRAIN| *** A_BUS| ***

Fit the Model with Only ASCs Discrete choice (multinomial logit) model Dependent variable Choice Log likelihood function Estimation based on N = 210, K = 3 Information Criteria: Normalization=1/N Normalized Unnormalized AIC Fin.Smpl.AIC Bayes IC Hannan Quinn R2=1-LogL/LogL* Log-L fncn R-sqrd R2Adj Constants only Response data are given as ind. choices Number of obs.= 210, skipped 0 obs Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] A_AIR| A_TRAIN| A_BUS| *** If the choice set varies across observations, this is the only way to obtain the restricted log likelihood.

Descriptive Statistics | Descriptive Statistics for Alternative AIR : | Utility Function | | 58.0 observs. | | Coefficient | All obs.|that chose AIR | | Name Value Variable | Mean Std. Dev.|Mean Std. Dev. | | | | | GC GC | | | | TTME TTME | | | | A_AIR ONE | | | | Descriptive Statistics for Alternative TRAIN : | Utility Function | | 63.0 observs. | | Coefficient | All obs.|that chose TRAIN | | Name Value Variable | Mean Std. Dev.|Mean Std. Dev. | | | | | GC GC | | | | TTME TTME | | | | A_TRAIN ONE | | |

Model Fit Based on Predictions  N j = actual number of choosers of “j.”  Nfit j =   i Predicted Probabilities for “j”  Cross tabulate: Predicted vs. Actual, cell prediction is cell probability Predicted vs. Actual, cell prediction is the cell with the largest probability N jk  =  i d ij  Predicted P(i,k)

Fit Measures Based on Crosstabulation | Cross tabulation of actual choice vs. predicted P(j) | | Row indicator is actual, column is predicted. | | Predicted total is F(k,j,i)=Sum(i=1,...,N) P(k,j,i). | | Column totals may be subject to rounding error. | NLOGIT Cross Tabulation for 4 outcome Multinomial Choice Model AIR TRAIN BUS CAR Total AIR | 32 | 8 | 5 | 13 | 58 | TRAIN | 8 | 37 | 5 | 14 | 63 | BUS | 3 | 5 | 15 | 6 | 30 | CAR | 15 | 13 | 6 | 26 | 59 | Total | 58 | 63 | 30 | 59 | 210 | NLOGIT Cross Tabulation for 4 outcome Constants Only Choice Model AIR TRAIN BUS CAR Total AIR | 16 | 17 | 8 | 16 | 58 | TRAIN | 17 | 19 | 9 | 18 | 63 | BUS | 8 | 9 | 4 | 8 | 30 | CAR | 16 | 18 | 8 | 17 | 59 | Total | 58 | 63 | 30 | 59 | 210 |

Using the Most Probable Cell | Cross tabulation of actual y(ij) vs. predicted y(ij) | | Row indicator is actual, column is predicted. | | Predicted total is N(k,j,i)=Sum(i=1,...,N) Y(k,j,i). | | Predicted y(ij)=1 is the j with largest probability. | NLOGIT Cross Tabulation for 4 outcome Multinomial Choice Model AIR TRAIN BUS CAR Total AIR | 40 | 3 | 0 | 15 | 58 | TRAIN | 4 | 45 | 0 | 14 | 63 | BUS | 0 | 3 | 23 | 4 | 30 | CAR | 7 | 14 | 0 | 38 | 59 | Total | 51 | 65 | 23 | 71 | 210 | NLOGIT Cross Tabulation for 4 outcome Multinomial Choice Model AIR TRAIN BUS CAR Total AIR | 0 | 58 | 0 | 0 | 58 | TRAIN | 0 | 63 | 0 | 0 | 63 | BUS | 0 | 30 | 0 | 0 | 30 | CAR | 0 | 59 | 0 | 0 | 59 | Total | 0 | 210 | 0 | 0 | 210 |

Effects of Changes in Attributes on Probabilities

Elasticities for CLOGIT Own effect Cross effects | Elasticity averaged over observations.| | Attribute is INVT in choice AIR | | Mean St.Dev | | * Choice=AIR | | Choice=TRAIN | | Choice=BUS | | Choice=CAR | | Attribute is INVT in choice TRAIN | | Choice=AIR | | * Choice=TRAIN | | Choice=BUS | | Choice=CAR | | Attribute is INVT in choice BUS | | Choice=AIR | | Choice=TRAIN | | * Choice=BUS | | Choice=CAR | | Attribute is INVT in choice CAR | | Choice=AIR | | Choice=TRAIN | | Choice=BUS | | * Choice=CAR | | Effects on probabilities of all choices in model: | | * = Direct Elasticity effect of the attribute. | Note the effect of IIA on the cross effects. Elasticities are computed for each observation; the mean and standard deviation are then computed across the sample observations.

Analyzing Behavior of Market Shares to Examine Discrete Effects  Scenario: What happens to the number of people who make specific choices if a particular attribute changes in a specified way?  Fit the model first, then using the identical model setup, add ; Simulation = list of choices to be analyzed ; Scenario = Attribute (in choices) = type of change  For the CLOGIT application ; Simulation = * ? This is ALL choices ; Scenario: GC(car)=[*]1.25$ Car_GC rises by 25%

Model Simulation | Discrete Choice (One Level) Model | | Model Simulation Using Previous Estimates | | Number of observations 210 | |Simulations of Probability Model | |Model: Discrete Choice (One Level) Model | |Simulated choice set may be a subset of the choices. | |Number of individuals is the probability times the | |number of observations in the simulated sample. | |Column totals may be affected by rounding error. | |The model used was simulated with 210 observations.| Specification of scenario 1 is: Attribute Alternatives affected Change type Value GC CAR Scale base by value The simulator located 209 observations for this scenario. Simulated Probabilities (shares) for this scenario: |Choice | Base | Scenario | Scenario - Base | | |%Share Number |%Share Number |ChgShare ChgNumber| |AIR | | | 1.973% 4 | |TRAIN | | | 1.748% 4 | |BUS | | |.903% 2 | |CAR | | | % -10 | |Total | | |.000% 0 | Changes in the predicted market shares when GC_CAR increases by 25%.

More Complicated Model Simulation In vehicle cost of CAR falls by 10% Market is limited to ground (Train, Bus, Car) CLOGIT ; Lhs = Mode ; Choices = Air,Train,Bus,Car ; Rhs = TTME,INVC,INVT,GC ; Rh2 = One,Hinc ; Simulation = TRAIN,BUS,CAR ; Scenario: GC(car)=[*].9$

Model Estimation Step Discrete choice (multinomial logit) model Dependent variable Choice Log likelihood function Estimation based on N = 210, K = 10 R2=1-LogL/LogL* Log-L fncn R-sqrd R2Adj Constants only Chi-squared[ 7] = Prob [ chi squared > value ] = Response data are given as ind. choices Number of obs.= 210, skipped 0 obs Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] TTME| *** INVC| *** INVT| *** GC|.07578*** A_AIR| *** AIR_HIN1| A_TRAIN| *** TRA_HIN2| *** A_BUS| *** BUS_HIN3| Alternative specific constants and interactions of ASCs and Household Income

Model Simulation Step | Discrete Choice (One Level) Model | | Model Simulation Using Previous Estimates | | Number of observations 210 | |Simulations of Probability Model | |Model: Discrete Choice (One Level) Model | |Simulated choice set may be a subset of the choices. | |Number of individuals is the probability times the | |number of observations in the simulated sample. | |Column totals may be affected by rounding error. | |The model used was simulated with 210 observations.| Specification of scenario 1 is: Attribute Alternatives affected Change type Value INVC CAR Scale base by value The simulator located 210 observations for this scenario. Simulated Probabilities (shares) for this scenario: |Choice | Base | Scenario | Scenario - Base | | |%Share Number |%Share Number |ChgShare ChgNumber| |TRAIN | | | % -3 | |BUS | | | % -3 | |CAR | | | 2.632% 6 | |Total | | |.000% 0 |

Compound Scenario: GC(Car) falls by 10%, TTME (Air,Train) rises by 25% NLOGIT ; Lhs = Mode ; Choices = Air,Train,Bus,Car ; Rhs = TTME,INVC,INVT,GC ; Rh2 = One,Hinc ; Simulation = AIR,TRAIN,BUS,CAR ; Scenario: invc(car)=[*].9 / ttme(air,train)=[*]1.25 $

Compound Scenario: GC(Car) falls by 10%, TTME (Air,Train) rises by 25% (at the same time) |Simulations of Probability Model | |Model: Discrete Choice (One Level) Model | |Simulated choice set may be a subset of the choices. | |Number of individuals is the probability times the | |number of observations in the simulated sample. | |Column totals may be affected by rounding error. | |The model used was simulated with 210 observations.| Specification of scenario 1 is: Attribute Alternatives affected Change type Value INVC CAR Scale base by value.900 TTME AIR TRAIN Scale base by value The simulator located 210 observations for this scenario. Simulated Probabilities (shares) for this scenario: |Choice | Base | Scenario | Scenario - Base | | |%Share Number |%Share Number |ChgShare ChgNumber| |AIR | | | % -23 | |TRAIN | | | % -15 | |BUS | | | 4.209% 9 | |CAR | | | % 29 | |Total | | |.000% 0 |

Willingness to Pay U(alt) = a j + b INCOME *INCOME + b Attribute *Attribute + … WTP = MU(Attribute)/MU(Income) When MU(Income) is not available, an approximation often used is –MU(Cost). U(Air,Train,Bus,Car) = α alt + β cost Cost + β INVT INVT + β TTME TTME + ε alt WTP for less in vehicle time = -β INVT / β COST WTP for less terminal time = -β TIME / β COST

WTP from CLOGIT Model Discrete choice (multinomial logit) model Dependent variable Choice Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] GC| INVT| *** TTME| *** AASC| *** TASC| *** BASC| *** WALD ; fn1=WTP_INVT=b_invt/b_gc ; fn2=WTP_TTME=b_ttme/b_gc$ WALD procedure Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] WTP_INVT| WTP_TTME|

Estimation in WTP Space

Nonlinear Utility Functions

Assessing Prospect Theoretic Functional Forms and Risk in a Non-linear Logit Framework: Valuing Reliability Embedded Travel Time Savings David Hensher The University of Sydney, ITLS William Greene Stern School of Business, New York University 8 th Annual Advances in Econometrics Conference Louisiana State University Baton Rouge, LA November 6-8, 2009

Prospect Theory  Marginal value function for an attribute (outcome) v(x m ) = subjective value of attribute  Decision weight w(p m ) = impact of a probability on utility of a prospect  Value function V(x m,p m ) = v(x m )w(p m ) = value of a prospect that delivers outcome x m with probability p m  We explore functional forms for w(p m ) with implications for decisions

Value and Weighting Functions

Choice Model U(j) = β ref + β cost Cost + β Age Age + β Toll TollASC + β curr w(p curr )v(t curr ) + β late w(p late ) v(t late ) + β early w(p early )v(t early ) + ε j Constraint: β curr = β late = β early U(j) = β ref + β cost Cost + β Age Age + β Toll TollASC + β[ w(p curr )v(t curr ) + w(p late )v(t late ) + w(p early )v(t early )] + ε j

Stated Choice Survey  Trip Attributes in Stated Choice Design Routes A and B Free flow travel time Slowed down travel time Stop/start/crawling travel time Minutes arriving earlier than expected Minutes arriving later than expected Probability of arriving earlier than expected Probability of arriving at the time expected Probability of arriving later than expected Running cost Toll Cost  Demographics: Age, Income, Gender

Survey Instrument

Data

Estimation Results

Choice Based Sampling  Over/Underrepresenting alternatives in the data set  May cause biases in parameter estimates. (Possibly constants only)  Certainly causes biases in estimated variances Weighted log likelihood, weight =  j / F j for all i. Fixup of covariance matrix – use “sandwich” estimator.  ; Choices = list of names / list of true proportions $ Choice`AirTrainBusCar True Sample

Choice Based Sampling Estimators Variable| Coefficient Standard Error b/St.Er. P[|Z|>z] Unweighted TTME| *** INVC| *** INVT| *** GC|.07578*** A_AIR| *** AIR_HIN1| A_TRAIN| *** TRA_HIN2| *** A_BUS| *** BUS_HIN3| Weighted TTME| *** INVC| *** INVT| *** GC|.10225*** A_AIR| *** AIR_HIN1| A_TRAIN| *** TRA_HIN2| *** A_BUS| *** BUS_HIN3|

Changes in Estimated Elasticities | Elasticity averaged over observations.| | Attribute is INVC in choice CAR | | Effects on probabilities of all choices in model: | | * = Direct Elasticity effect of the attribute. | | Unweighted | | Mean St.Dev | | Choice=AIR | | Choice=TRAIN | | Choice=BUS | | * Choice=CAR | | Weighted | | Mean St.Dev | | Choice=AIR | | Choice=TRAIN | | Choice=BUS | | * Choice=CAR |

The I.I.D Assumption U itj =  ij +  ’x itj +  ’z it +  ijt F(  itj ) = 1 – Exp(-Exp(  itj )) (random part of each utility) Independence across utility functions Identical variances (means absorbed in constants) Restriction on equal scaling may be inappropriate Correlation across alternatives may be suppressed Equal cross elasticities is a substantive restriction Behavioral implication of independence from irrelevant alternatives is unreasonable (IIA). If an alternative is removed, probability is spread equally across the remaining alternatives.

A Hausman Test for IIA  Estimate full model with “irrelevant alternatives”  Estimate the short model eliminating the irrelevant alternatives Eliminate individuals who chose the irrelevant alternatives Drop attributes that are constant in the surviving choice set.  Do the coefficients change? Use a Hausman test: Chi-squared, d.f. Number of parameters estimated  Practicalities: Fit the model, then again with ;IAS = the irrelevant alternative(s)

IIA Test /* Using the internal routine usually means specifying */ an unreasonable model. It is also easy to program directly. clogit;lhs=mode;choices=air,train,bus,car ;rhs=gc,ttme,invc,invt,aasc,tasc,basc$ matrix;bfull=b(1:4);vfull=varb(1:4,1:4)$ create ; j = trn(-4,0)$ reject ; j=1 | chair=1 $ clogit;lhs=mode;choices=train,bus,car ;rhs=gc,ttme,invc,invt,tasc,basc$ matrix;bshort=b(1:4);vshort=varb(1:4,1:4)$ matrix;d=bshort-bfull;v=vshort-vfull$ matrix;list;iiatest=d' d$ calc;list;ctb(.95,4)$

IIA Test for Choice AIR |Variable| Coefficient | Standard Error |b/St.Er.|P[|Z|>z]| GC | TTME | INVC | INVT | AASC | TASC | BASC | GC | TTME | INVC | INVT | TASC | BASC | Matrix IIATEST has 1 rows and 1 columns | Test statistic | Listed Calculator Results | Result = Critical value

Case Study – Omitted Attributes  Do all consumers evaluate all attributes? An information processing strategy – minimize processing cost Lexicographic preferences – some attributes are irrelevant.  Do we know which attributes are evaluated?  How to incorporate omitted attributes information in the model Zero fill in the data? Zero is not a valid PRICE. Change the equation – True zeros in utility functions Some consumers do not “value” some attributes.

Modeling Attribute Choice  Conventional: U j =  ′x j. For ignored attributes, set x k,ijt =0. This eliminates x kj from the utility function Price = 0 is not a reasonable datum. Distorts choice probabilities  Appropriate: Formally set  k = 0 Requires a ‘person specific’ model Accommodate as part of model estimation

Choice Strategy Heterogeneity  Methodologically, a rather minor point – construct appropriate likelihood given known information  Not a latent class model. Classes are not latent.  Not the ‘variable selection’ issue (the worst form of “stepwise” modeling)  Familiar strategy gives the wrong answer.

Application: Sydney Commuters’ Route Choice  Stated Preference study – several possible choice situations considered by each person  Multinomial and mixed (random parameters) logit  Consumers included data on which attributes were ignored.  (Ignored attributes coded -888 in NLOGIT are automatically treated by constraining β=0 for that observation.)

Data for Application of Information Strategy S tated/Revealed preference study, Sydney car commuters surveyed, about 10 choice situations for each. E xisting route vs. 3 proposed alternatives. A ttribute design Original: respondents presented with 3, 4, 5, or 6 attributes Attributes – four level design.  Free flow time  Slowed down time  Stop/start time  Trip time variability  Toll cost  Running cost Final: respondents use only some attributes and indicate when surveyed which ones they ignored

Discrete Choice Model Extensions Heteroscedasticity and other forms of heterogeneity Across individuals Across alternatives Panel data (Repeated measures) Random and fixed effects models Building into a multinomial logit model The nested logit model Latent class model Mixed logit, error components and multinomial probit models A Generalized Mixed Logit Model – The frontier Combining revealed and stated preference data